In [1]:
from IPython.core.display import display, HTML
display(HTML("""<style> .container {width:96% !important;}</style>"""))
from IPython.display import IFrame
In [2]:
import pandas as pd
import numpy as np
# from plotly.offline import init_notebook_mode, iplot
# import cufflinks as cf
# init_notebook_mode()
# cf.go_offline()
from __future__ import division
In [3]:
!ls additional_data
path = 'additional_data/'
In [4]:
nat34 = pd.read_csv(path + 'nat03_04.csv')
nat56 = pd.read_csv(path + 'nat05_06.csv')
nat78 = pd.read_csv(path + 'nat07_08.csv')
In [5]:
print nat34.shape
print nat56.shape
print nat78.shape
In [6]:
nat34.head().T
Out[6]:
In [7]:
# Default if MIS_status = CHGOFF, else MIS_status = PIF
print nat34.MIS_Status.value_counts()
print 'Default rate =', (nat34.MIS_Status == 'CHGOFF').sum() / len(nat34)
In [8]:
print (nat34.MIS_Status == 'CHGOFF').sum() / len(nat34)
print (nat56.MIS_Status == 'CHGOFF').sum() / len(nat56)
print (nat78.MIS_Status == 'CHGOFF').sum() / len(nat78)
In [9]:
nat34.BalanceGross.value_counts()
Out[9]:
EDA on features¶
In [10]:
nat34.State.value_counts().head()
Out[10]:
In [11]:
nat34.City.value_counts().head()
Out[11]:
In [12]:
nat34.Zip.value_counts().head()
Out[12]:
In [13]:
nat34.UrbanRural.value_counts()
Out[13]:
In [14]:
nat34.Bank.value_counts().head()
Out[14]:
In [15]:
nat34.NAICS.value_counts().head()
Out[15]:
In [16]:
nat34['ApprovalDate'] = pd.to_datetime(nat34.ApprovalDate)
In [17]:
nat34.ApprovalDate.dt.month.value_counts()
Out[17]:
In [18]:
nat34.ApprovalFY.value_counts().head()
Out[18]:
In [19]:
nat34.NoEmp.value_counts().head()
Out[19]:
In [20]:
nat34.NewExist.value_counts()
Out[20]:
In [21]:
nat34[nat34.MIS_Status == 'CHGOFF'].sample(10).T
Out[21]:
In [ ]:
nat34[nat34.MIS_Status == 'CHGOFF'].City.value_counts().head(10).iplot(kind = 'bar')
In [29]:
nat34[nat34.MIS_Status == 'CHGOFF'].Zip.value_counts().head(10)
Out[29]:
In [32]:
nat34[nat34.MIS_Status == 'CHGOFF'].Zip.astype(str).str[:3].value_counts().head(10)
Out[32]:
In [35]:
nat34[nat34.MIS_Status == 'CHGOFF'].City.astype(str).str[:3].value_counts().head(10)
Out[35]:
In [37]:
nat34[nat34.MIS_Status == 'CHGOFF'].Bank.value_counts().head()
Out[37]:
In [38]:
nat34.UrbanRural.value_counts()
Out[38]:
In [39]:
nat34.RevLineCr.value_counts()
Out[39]:
In [ ]:
train[train.default == 1].Term.iplot(kind = 'hist')
In [ ]:
train[train.default == 0].Term.iplot(kind = 'hist')